The highly anticipated Nvidia OpenAI investment, once touted as a massive $100 billion deal, is now shrouded in uncertainty after months of silence. The initial letter of intent has not materialized into a solid agreement, raising significant questions in the tech industry about the future of the partnership between the two AI giants.
What Happened to the Nvidia OpenAI Investment?
It all began in the fall of 2025 when Nvidia and OpenAI announced a letter of intent for an investment of up to $100 billion to enhance AI infrastructure. According to statements at the time, the plan involved installing 10 gigawatts of Nvidia systems for OpenAI. This was an incredibly ambitious project, equivalent to the power generated by approximately 10 nuclear reactors. Nvidia CEO Jensen Huang had even stated, “This is a giant project,” noting it would match the company’s total GPU shipments for that year.
However, a crucial detail from the start was that the announcement was a non-binding “letter of intent,” not a legally enforceable contract. In recent weeks, Nvidia CEO Jensen Huang has downplayed this colossal figure. Huang clarified that the $100 billion “was never a commitment” and that OpenAI had invited them to invest “up to” that amount. While he affirmed Nvidia would definitely be involved in what would likely be the largest investment in the company’s history, he made it clear the final number would not reach the $100 billion level.
In response to these developments, OpenAI CEO Sam Altman attempted to smooth things over with a post on X, stating: “We love working with NVIDIA, and they make the best AI chips in the world. We hope to be a giant customer of theirs for a very long time. I don’t understand where all this craziness is coming from.”
OpenAI’s Search for Alternatives and the Chip Wars
One of the primary reasons behind the deal’s uncertainty is rumored to be OpenAI’s dissatisfaction with the performance of Nvidia chips, particularly in “inference” tasks. Inference is the process where a trained AI model generates responses to user queries. According to reports, this performance issue became especially apparent with OpenAI’s code-generation tool, Codex. Consequently, this situation has pushed OpenAI to seek alternatives to reduce its dependency on Nvidia.
OpenAI’s steps in this search have been quite noteworthy:
- Cerebras Agreement: In January, OpenAI announced a $10 billion deal with Cerebras, a company that produces specialized chips to reduce inference latency.
- AMD Partnership: The company signed an agreement with AMD in October for the supply of six gigawatts of GPUs.
- Custom Chip Development: Furthermore, OpenAI revealed plans to develop a custom AI chip with Broadcom to completely break its reliance on Nvidia.
Adding another layer to the competition, Groq, another startup OpenAI was in talks with, was acquired by Nvidia through a $20 billion licensing deal, showcasing the fierce rivalry in the AI market. These events have also reignited criticisms of “cyclical investing” in the tech world. Critics argue that Nvidia has created a system where it invests in startups, which then use that money to purchase GPUs from Nvidia. The failure of the $100 billion plan to materialize highlights questions about the sustainability of this investment model and suggests a potential reshaping of power dynamics in the AI chip market.
So, what are your thoughts on the investment uncertainty between Nvidia and OpenAI? Share your opinions with us in the comments!
{{user}} {{datetime}}
{{text}}